Bug in drawing WMS-layer with mapscript since fix of Bug 541

[silke@…]

It seems that the fix for Bug 541 has introduced a new one which is only
existent when drawing WMS-layers with mapscript (at least for python mapscript):

See following python script:

import mapscript
theMap = mapscript.mapObj('/home/silke/testwms.map')
theMap.draw()

with mapserver-4.4.0_beta2 this works OK. With mapserver 4.4.1 it doesn't work
any more, giving the following error::
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.2/site-packages/mapscript.py", line 1419, in draw
    def draw(*args): return _mapscript.mapObj_draw(*args)
_mapscript.MapServerError: msRASTERLayerOpen(): General error message. Attempt o
open a RASTER layer, but this is only supported after a raster query.

Using mapserver CGI is OK with both versions. Using the debug mode and
introducing further debug statements I tripped it down to the function which has
been changed in 541. Assefa, I put you in CC since you fixed this bug. 

Before I go into further debugging I wanted to ask whether someone has further
ideas.

See below the mapfile I used to reproduce the reported bug.

Greetings,

          Silke


---------------------
NAME "TESTWMS"
STATUS ON
SIZE 600 500
EXTENT  3427000 5787000 3444200 5801000
UNITS METERS

PROJECTION
        # ETRS89 / UTM  zone 33N = 15. Laengengrad
         #+proj=utm
         #+zone=33
         #+ellps=GRS80
         #+units=m
         #+no_defs
        "init=epsg:31463"
END

OUTPUTFORMAT
  NAME png
  DRIVER "GD/PNG"
  MIMETYPE "image/png"
  IMAGEMODE RGB
  EXTENSION "png"
END

# WEB Objekt
WEB
        IMAGEPATH "/tmp/"
        IMAGEURL "/tmp/"
END

LAYER
        NAME "frida"
        CONNECTIONTYPE WMS
        CONNECTION "http://demo.intevation.de/cgi/frida-wms"
        METADATA
                "wms_srs" "EPSG:31463"
                "wms_name"    "gruenflaechen"
                "wms_server_version"  "1.1.1"
                "wms_format"          "image/png"
        END
        TYPE RASTER
        STATUS ON
END

END

--------------------

[assefa]

I am not 100% sure but It seems for me that this is related to bug 1118. It 
has been fixed in mapserver 4.4.2 and in current CVS version. It certainly was 
not fixed in 4.4.1.
If possible could you please use the 4.4.2 version and let me know if it 
reproduces with it. If it does I could try to track it.

[silke@…]

Yes, you are right. Mapserver 4.4.2 works well. Sorry for bothering you.

Many thanks,
         
         Silke